Neuron-specific WWOX gene therapy produces dose-dependent, durable rescue in a model of WWOX-related epileptic encephalopathy Authors: Mustafa Obeid, Rania Akkawi, Srinivasarao Repudi, Prince Kumar Singh, Baraa Abudiab, Takwa Jebara, Allyson Berent, Thomas Brennan, Yael Weiss, Tawfeeq Shekh-Ahmad, Rami I Aqeilan Published: 2026 ABSTRACT AbstractBi-allelic WWOX loss-of-function mutations underlie a spectrum of developmental and epileptic encephalopathies, including the severe WOREE syndrome and the milder SCAR12, for which no effective therapies exist. Gene replacement represents a compelling strategy; however, critical parameters for clinically translatable delivery remain undefined. Here, we establish a rationally optimized WWOX gene therapy framework using a severe Wwox-null mouse model. Systematic evaluation of promoter selection, cellular targeting, and vector configuration, dose, and timing identified neuron-restricted expression driven by the human synapsin I promoter as the most effective approach, yielding robust and durable phenotypic rescue compared to non-specific or oligodendrocyte-directed strategies. Removal of the WPRE element enabled controlled transgene expression and facilitated dose calibration within a clinically relevant range. An optimal AAV9-hSynI-WWOX dose restored survival, growth, metabolic function, behavior, and fertility to near wild-type levels. Therapeutic efficacy was associated with sustained reconstitution of WWOX across central and peripheral neural tissues, absence of off-target hepatic expression, enhanced myelination, and reduced neuroinflammation. Early neuronal hyperexcitability was suppressed following treatment. Importantly, efficacy was shown in an early postnatal therapeutic window (P1–P5), supporting durable rescue. Collectively, these findings define key design and dosing principles for translation of WWOX gene therapy to patients with developmental and epileptic encephalopathies.Keywords: gene therapy, WOREE syndrome, DEE, WWOX, WPRE Introduction The WW domain-containing oxidoreductase (WWOX) gene spans one of the most active common fragile sites. Common fragile sites have been suggested to play roles in a wide range of neurodevelopmental processes, including cell-cell adhesion, synapse organization, interneuron differentiation, neuronal projection development, glutamate receptor signaling, and CNS axonogenesis.1 WWOX maps to chromosome 16q23.3–24.1 and encodes a 46 kDa protein composed of 414 amino acids, with key roles in tumor suppression, DNA damage response, and metabolic regulation.2,3 Structurally, WWOX features two N-terminal WW domains, a nuclear localization signal (NLS), and a C-terminal short-chain dehydrogenase/reductase (SDR) domain, enabling it to interact with multiple protein partners and engage in diverse signaling pathways.4,5,6,7 WWOX has been extensively studied for its tumor suppressor functions, primarily through regulation of key cancer-associated signaling pathways, including WNT, TGFβ1, NF-κB, and receptor tyrosine kinases (RTKs).8,9,10,11,12 In addition, WWOX modulates cellular metabolism via HIF1α and IDH pathways13,14 and contributes to the DNA damage response through interactions with ATM,14 BRCA1,15,16 and p73.17,18 These activities highlight its key role in tumor formation. More recently, increasing evidence indicates a vital role for WWOX in the nervous system. WWOX is highly expressed in the brain, shows dynamic patterns during embryonic and postnatal development, and is crucial for maintaining the central nervous system (CNS) homeostasis.8,19,20,21,22,23 Mutations in WWOX have been linked to a broad spectrum of neurological disorders, including Alzheimer’s disease,24,25,26 Parkinson’s disease,27,28 multiple sclerosis,29,30,31 autism spectrum disorders,32,33 epilepsy, and ataxia,34 underscoring its essential role in CNS health and function. Of particular interest, germline recessive mutations including missense, nonsense, and deletion variants are causally associated with two major inherited syndromes: SCAR12 (spinocerebellar ataxia, autosomal recessive 12; MIM 614322) and WOREE syndrome (WWOX-related epileptic encephalopathy, also classified as developmental and epileptic encephalopathy 28; MIM 616211). WOREE syndrome is a severe, early-onset neurological disorder caused primarily by bi-allelic mutations that result in premature stop codons or complete loss of WWOX expression. Clinical features include profound global developmental delay and intractable epilepsy, with seizure types such as tonic, clonic, tonic-clonic, myoclonic, infantile spasms, and absence seizures. Most affected individuals fail to reach basic developmental milestones, including eye contact, sitting, speech, or ambulation. Approximately 50% of patients with WOREE syndrome harboring null mutations die by 5 years of age. In contrast, SCAR12 presents with a milder phenotype, often involving early-onset epilepsy and cerebellar ataxia that may respond to anti-epileptic drugs (AEDs), although patients frequently experience intellectual disability and motor dysfunction. WWOX mutations have also been implicated in West syndrome, which is characterized by infantile spasms and hypsarrhythmia. Magnetic resonance imaging (MRI) of children with WWOX mutations commonly reveals structural abnormalities such as corpus callosum (CC) hypoplasia, delayed myelination, progressive cerebral atrophy, and optic nerve atrophy.35,36,37,38,39,40,41 Previous work from our laboratory sought to define the cell populations underlying the WWOX-deficient phenotype. Using conditional mouse models, we showed that deletion of Wwox in neural stem and progenitor cells driven by the nestin promoter (N-KO), or in postmitotic neurons driven by the synapsin promoter (S-KO), leads to severe epilepsy, hypomyelination, ataxia, and early lethality by 3–4 weeks of age, closely phenocopying global Wwox-null mice.42 These observations established neuronal WWOX as indispensable for normal brain function and survival and provided the rationale for testing whether targeted restoration of WWOX expression could reverse these profound neurological deficits.43,44 In our prior proof-of-concept study, we demonstrated that neuron-targeted Adeno-associated virus 9 (AAV9)-mediated delivery of WWOX achieves widespread neuronal expression in the developing brain and ameliorates disease manifestations in Wwox-deficient mice.45 However, key gaps regarding optimal dosing and safety remained, thus prompting the present study. In this study, we refine the WWOX gene therapy approach to improve effectiveness for potential clinical use. By driving WWOX expression under the neuron-specific synapsin I promoter, we achieve targeted and strong neuronal transduction. To better regulate transgene expression, we generated a vector lacking the woodchuck hepatitis virus post-transcriptional regulatory element (WPRE) element. Additionally, we demonstrated a dose-dependent therapeutic effect, with neuronal WWOX restoration in Wwox-null mice rescuing key phenotypes, including premature death, growth delay, hypoglycemia, and behavioral deficits to levels indistinguishable from WT mice. Furthermore, this approach significantly enhances CNS myelination, decreases neuroinflammation, and lessens gliosis. These findings, along with growing evidence linking WWOX mutations to various neurological disorders, highlight the potential of WWOX gene replacement therapy. Overall, our results provide a solid foundation for developing effective treatments for WOREE syndrome, SCAR12, and likely other WWOX-related neurodevelopmental and neurodegenerative diseases. Results Comparative promoter analysis implicates neuron-specific WWOX expression via the synapsin promoter in improved disease rescue Given the essential role of neuronal WWOX in CNS homeostasis, we systematically evaluated AAV-mediated human WWOX (hWWOX) expression using promoters selected to test expression breadth, strength, and cell-type specificity. Ubiquitous promoters (EF1α, CMV, and CBA) were used to assess whether broad or strong expression could compensate for WWOX loss, whereas the myelin basic protein (MBP) promoter tested oligodendrocyte-restricted rescue and the synapsin I promoter enabled neuron-specific restoration. Therapeutic efficacy was evaluated by survival, body weight, and blood glucose levels. EF1α, a broadly active but relatively weak promoter,46 drove widespread WWOX expression but failed to rescue lethality, growth retardation, or hypoglycemia in Wwox-null mice (Figures 1A–1C and 1M; Figures S1A and S1B; S2D). Stronger ubiquitous promoters (CMV and CBA),47,48 supported early postnatal rescue of survival and glycemic defects, indicating that early high-level expression is sufficient for transient benefit, but this protection was not sustained at later stages (Figures 1D, 1F, and 1M; Figures S1A and S1B; Figures S2A–S2D). In contrast, oligodendrocyte-specific WWOX expression driven by the MBP promoter resulted in weak expression and no phenotypic improvement, demonstrating that restoration in this lineage alone is insufficient (Figures 1G–1I and 1M; Figures S1A and S1B; Figure S2D). WWOX restoration driven by the synapsin promoter achieves high expression efficiency (A, D, G, and J) Kaplan-Meier survival analyses of Wwox-null mice treated with AAV9-EF1a-hWWOX-WPRE (4 × 1010 vg, n = 5), AAV9-CMV- hWWOX -WPRE (4 × 1010 vg, n = 5), AAV9-MBP- hWWOX -WPRE (4 × 1010 vg, n = 5), and AAV9-hSynI- hWWOX -WPRE (4 × 1010 vg, n = 6), respectively, compared with Wwox-null (KO) and wild type (WT) mice injected with the reference item (RI) (n = 6). p < 0.01, log rank (Mantel-Cox) test. (B, E, H, and K) Graphs showing body weight (g) of WT mice (n = 6), KO mice injected with the RI (n = 6), and KO mice injected with hWWOX driven by different promoters (n = 5–6) at multiple postnatal time points. At postnatal day 14 (P14), KO mice treated with hWWOX-hSynI and hWWOX-CMV exhibited significant increases in body weight compared with RI-injected KO mice (p < 0.001 and p < 0.01, respectively), whereas no statistically significant differences were observed in KO mice treated with hWWOX-EF1α or hWWOX-MBP. (C, F, I, and L) Bar graphs representing glucose levels (mg/dL) at P14 in WT mice (n = 6), KO mice injected with RI (n = 6), and KO mice injected with hWWOX using different promoters (n = 5–6). p < 0.001 for KO mice injected with hWWOX-hSynI and hWWOX-CMV, but no significant difference was observed for KO mice injected with hWWOX-Ef1a and KO mice injected with hWWOX-MBP compared with KO mice injected with RI. (M) Immunohistochemistry staining for WWOX (red) in the cortex of P18 mice indicating the specificity and intensity of WWOX restoration using different promoters. Strikingly, neuron-specific expression under the human synapsin I (hSynI) promoter yielded the most robust and sustained rescue, with efficient neuronal expression in vivo (Figures 1J–1M; Figures S1A–S1D; Figure S2D) and superior neuron-restricted transduction in primary neuronal cultures compared with CMV, while MBP-driven vectors showed no neuronal expression (Figure S2E). These findings align with our earlier conditional knockout (KO) models.42 Moreover, the comparative promoter analysis indicates that neuron-specific WWOX expression driven by the synapsin promoter is associated with improved disease rescue. While transgene expression was not quantitatively normalized across promoter conditions, the observed differences likely reflect a combination of cell-type specificity and expression dynamics. Therefore, synapsin I promoter was used for all subsequent experiments. Removal of WPRE enables controlled WWOX expression while maintaining therapeutic efficacy at higher vector doses In our earlier therapeutic construct, we incorporated the WPRE, a widely used enhancer of transgene expression that increases mRNA stability, promotes nuclear export, and improves transcript termination, typically resulting in a 5- to 8-fold increase in expression.49 As expected, the presence of WPRE downstream of the transgene markedly augmented WWOX expression when compared with wild-type (WT) expression (Figure 2). To directly evaluate the intrinsic therapeutic efficacy of WWOX expression and enable tighter control of transgene output, we subsequently removed the WPRE element and assessed WWOX expression and the ability of WWOX alone to rescue phenotypes in Wwox-null mice. Consequences of WPRE removal on WWOX expression and therapeutic efficacy (A) Schematic representation of AAV9 vectors expressing WWOX under the human synapsin (hSynI) promoter, with or without the WPRE regulatory element. (B) Kaplan-Meier survival analyses of Wwox-null mice treated with AAV9-hSynI-hWWOX (4 × 1010 vg), AAV9-hSynI-hWWOX-WPRE (4 × 1010 vg), or AAV9-hSynI-hWWOX (8 × 1010 vg). p < 0.01, log rank (Mantel-Cox) test. The black dot indicates that this mouse was collected for analysis. (C) A graph showing body weight (g) monitoring of WT (n = 5), KO (n = 4), and KO mice injected with AAV9-hSynI-hWWOX (4 × 1010 vg, n = 3), AAV9-hSynI-hWWOX-WPRE (4 × 1010 vg, n = 3), or AAV9-hSynI-hWWOX (8 × 1010 vg, n = 5). (D) A graph showing glucose monitoring (mg/dL) of WT (n = 5), KO (n = 4), and KO mice injected with AAV9-hSynI-hWWOX (4 × 1010 vg,n = 3), AAV9-hSynI-hWWOX-WPRE (4 × 1010 vg, n = 3), or AAV9-hSynI-hWWOX (8 × 1010 vg, n = 5) (E) Western blot analysis of WWOX protein levels in cortex, hippocampus, midbrain, and cerebellum of WT and KO mice treated with AAV9-hSynI-hWWOX (W) or AAV9-hSynI-hWWOX-WPRE (W + WPRE) at P0. Quantification indicates WWOX intensity relative to WT. (F and G) Representative immunohistochemistry images showing WWOX (gray) and the neuronal marker NeuN (green) in the cortex and cerebellum, respectively, of WT, KO, and KO mice treated with AAV9-hSynI-hWWOX (W) or AAV9-hSynI-hWWOX-WPRE (W + WPRE). Hoechst (blue) labels nuclei. Scale bars, 50 μm. To directly assess the impact of WPRE on WWOX levels, we compared AAV-SynI-mediated WWOX expression with or without WPRE following intracerebroventricular (ICV) delivery at different doses (Figure 2A). At a dose of 4 × 1010 vg, exclusion of WPRE was insufficient to rescue lethality in KO mice, whereas treatment with WPRE-containing vectors at the same dose resulted in survival rates comparable to WT (Figure 2B). Consistently, KO mice treated with WPRE-lacking vectors exhibited growth retardation and reduced glucose levels relative to WT (Figures 2C and 2D). Increasing the dose of the WPRE-lacking vector to 8 × 1010 vg was associated with improved outcomes, including rescue of lethality and normalization of growth and glucose levels. Immunoblot and immunohistochemistry analyses revealed that WPRE significantly enhanced WWOX protein levels in a dose-dependent and regionally consistent manner across the cortex, hippocampus (Hippo), midbrain (Mid), and cerebellum (Cereb) (Figures 2E–2G; Figure S3A). Quantification of WWOX protein levels confirmed that inclusion of WPRE significantly increased expression at both tested doses (Figure S3B). Furthermore, quantification of WWOX immunohistochemical staining showed a significant increase in WWOX-positive NeuN+ neurons, rising from ∼40% to ∼55%–60% following treatment (p < 0.05) (Figure S3C). Notably, WWOX expression driven by the WPRE-containing vector appears more intense and diffuse, with a prominent punctate signal throughout neurons, precluding quantification of WWOX-positive neurons, which may reflect higher expression levels rather than a true change in subcellular localization. Furthermore, immunohistochemical and immunoblot analyses of WWOX expression confirmed a dose-dependent increase, with low expression observed at 4 × 1010 vg in the absence of WPRE and markedly higher levels at 8 × 1010 vg. Notably, inclusion of WPRE at 4 × 1010 vg resulted in substantially higher expression (Figures S3D–S3E). Interestingly, increasing the vector dose in the absence of WPRE failed to recapitulate the expression levels achieved with lower dose (LD) containing WPRE (Figure S3F). This may suggest that the non-WPRE vector yields a more uniform, predominantly cytoplasmic WWOX pattern that resembles endogenous expression in WT neurons (Figures 2F and 2G; Figure S3A). Taken together, these findings imply that removal of WPRE is associated with a more controlled neuronal WWOX expression profile, underscoring the importance of regulated transgene expression for translational gene therapy approaches. However, this reduction in expression necessitates the use of higher vector doses to achieve comparable therapeutic outcomes. WWOX gene therapy improves survival and systemic phenotypes in a dose-dependent manner Building on our optimization studies (Figure 2; Figure S3), we generated AAV9-hSynI-WWOX vectors lacking the WPRE element to achieve tighter control of transgene expression and to proactively optimize the safety margin for clinical translation. For translational relevance, we evaluated two clinically applicable doses: an LD (1.23 × 1011 vg) and a higher dose (HD, 2.63 × 1011 vg) (Figure 3A). Survival analysis revealed a clear dose-response relationship: LD treatment modestly, though significantly, extended lifespan relative to untreated Wwox-null controls (KO mice injected with the reference item (RI), whereas HD treatment produced a marked and sustained survival benefit (Figure 3B). WWOX gene therapy enhances survival in a dose-dependent manner (A) Schematic representation of the AAV9-hSynI-hWWOX vectors used for neonatal intracerebroventricular (ICV) injections. Two viral doses were tested: a low dose (LD, 1.23 × 1011 vg) and a high dose (HD, 2.63 × 1011 vg). (B) Kaplan-Meier survival analyses of Wwox-null mice treated with HD or LD AAV9-SynI-WWOX compared with KO and WT mice injected with the RI. HD produced survival comparable to WT animals (p = 0.78), and significantly extended survival relative to the LD group (p < 0.0001) and the KO + RI controls (p < 0.0001). LD also significantly improved survival compared with the KO + RI controls (p < 0.0001). (C) A graph representing body weight during the first postnatal month in KO + LD and KO + HD mice, with weights reaching near WT levels by postnatal day 30, whereas KO mice exhibited marked growth retardation. (D) Long-term growth monitoring demonstrated sustained weight gain and normalization in KO + HD mice up to 180 days of age. (E–G) Blood glucose levels (mg/dL) measured at postnatal days 10, 20, and 30. (H) A graph representing glucose measurements (mg/dL) of KO + HD mice at P180 was indistinguishable from that of WT mice. Statistical analysis was performed using Student’s t test (∗p < 0.05; ∗∗p < 0.01; ns, not significant). Error bars represent mean ± SD. Growth trajectories closely paralleled survival outcomes. Untreated Wwox-null mice exhibited severe growth retardation compared with WT controls, while both LD- and HD-treated groups showed progressive weight gain (Figure 3C). By the study endpoint, body weights of HD-treated mice were nearly indistinguishable from WT littermates (Figure 3D), with no significant sex-dependent differences observed. Given the profound hypoglycemia characteristic of Wwox-null mice,45 we next assessed glucose homeostasis longitudinally. Beginning at postnatal day 10 (P10), both LD and HD treatments partially corrected hypoglycemia. By P20, glucose levels in HD-treated mice were fully normalized and comparable to WT controls, whereas LD-treated mice showed an intermediate improvement. From P30 onward, glucose levels were normalized across all groups, with HD-treated mice maintaining stable WT-like glycemia throughout the study (Figures 3E–3H). Motor coordination was assessed at P18 using hindlimb clasping test.45 As shown in Figures S4A and S4B, untreated Wwox-null mice displayed severe motor impairment, whereas LD treatment resulted in partial improvement and HD treatment achieved near-complete rescue, with ataxia scores approaching those of WT controls. Finally, to determine whether neuronal WWOX restoration ameliorates broader systemic deficits, we assessed reproductive function. Fertility testing using 20 breeding cages per group demonstrated that HD-treated Wwox-null mice exhibited fertility rates and litter sizes comparable to WT and heterozygous controls (Figures S4C–S4E). Collectively, these data demonstrate that AAV9-hSynI-hWWOX gene therapy confers robust, dose-dependent benefits across survival, growth, glycemic regulation, motor coordination, and fertility. While both doses improved disease phenotypes, the HD consistently achieved near-complete rescue, establishing a strong therapeutic window and supporting the clinical potential of precisely tuned neuronal WWOX replacement. Neuronal WWOX restoration normalizes neurobehavioral function To determine whether neuronal WWOX reconstitution rescues behavioral and motor deficits, we assessed Wwox-null mice treated with high-dose AAV9-hSynI-WWOX (KO+WWOX HD) and compared them with WT controls (Figure 4A). Behavioral testing could not be performed in untreated Wwox-null mice due to severe morbidity and early lethality, and LD-treated mice did not survive to P90; therefore, analyses were limited to WT and HD-treated groups. High-dose neuronal WWOX expression restores neurobehavioral performance (A) Schematic representation of the experimental design. KO mice were injected at postnatal day 0–1 (P0‑P1) with AAV9-hSynI-hWWOX at HD (2.63 × 1011 vg) and monitored for survival, body weight, glucose levels, behavioral symptoms, and seizure activity. Behavioral assessments were performed at 3 months of age, including open field, elevated plus maze, and rotarod tests. (B) Representative locomotor tracking plots of male and female WT (n = 9) and KO + W HD (n = 10) mice in the open field test. (C) Representative trajectory maps from the elevated plus maze of male and female WT (n = 9) and KO + W HD groups (n = 10). (D–G) Quantitative analysis of open field parameters: velocity (cm/s), total distance traveled (cm), frequency in zone center, and frequency in periphery, respectively. (H–J) Quantitative analysis of elevated plus maze test parameters: velocity (cm/s), duration in the open arm (s), and duration in the closed arm (s), respectively. (K) A graph representing the latency (s) of the mice (WT n = 9, KO + W HD n = 10) to fall from the rotarod. Statistical analysis was performed using Student’s t test (∗p < 0.05; ns, not significant), Error bars represent mean ± SD. At three months of age, both male and female HD-treated mice displayed open-field performance indistinguishable from WT controls. Locomotor activity parameters, including movement velocity and total distance traveled, as well as spatial exploration of center and periphery zones, showed no significant differences between groups (Figures 4B, 4D–4G). In the elevated plus maze, HD-treated mice exhibited anxiety-related behaviors comparable to WT mice, indicating normalization of affective responses (Figures 4C, 4H–4J). Treated KO mice exhibited significantly higher motor coordination and learning compared with WT mice (Figure 4K). Together, these findings indicate that early, high-dose neuronal WWOX gene therapy yielded neurobehavioral and motor outcomes indistinguishable from those of WT controls, encompassing locomotor activity, anxiety-related behavior, and motor coordination. AAV9-SynI-WWOX restores WWOX DNA, mRNA, and protein expression in a dose-dependent and durable manner To assess the efficiency and biodistribution of AAV9-mediated WWOX delivery, we quantified viral DNA (vDNA) levels in brain tissue at P30 using qPCR. Analysis of the cortex, hippocampus, midbrain, and cerebellum revealed a clear dose-dependent pattern, with higher vDNA levels in HD-treated mice than LD-treated mice across regions, with statistical significance in the hippocampus (Figures 5A–5D), confirming efficient transduction and dose-dependent vector distribution. AAV9-hSynI-hWWOX increases WWOX DNA, mRNA, and protein expression in Wwox-null mice in a dose-dependent manner (A–D) Vector genome copies (GC) in cortex, hippocampus, midbrain, and cerebellum at postnatal day 30 (P30), measured by qPCR and normalized to WT levels. (E–H) WWOX mRNA expression in the same brain regions, normalized to WT. (I–L) Representative western blot analysis showing WWOX protein expression in cortex, hippocampus, midbrain, and cerebellum, respectively, following LD (1.23 × 1011 vg) and HD (2.63 × 1011 vg) AAV9-hSynI-hWWOX treatments. Quantification indicates WWOX intensity relative to WT. Statistical analysis was performed using Student’s t test (∗p < 0.05; ns, not significant; in G and H it is near significant, p value = 0.08 and 0.06, respectively). Error bars represent mean ± SD. To determine whether the delivered transgene was transcriptionally active, we next measured WWOX mRNA levels. RT-qPCR analysis demonstrated robust, region-wide induction of WWOX transcripts in HD-treated mice, whereas LD-treated mice showed lower but readily detectable expression (Figures 5E–5H). These data establish that AAV9-hSynI-hWWOX drives efficient, dose-dependent neuronal transcription of the therapeutic transgene. We then evaluated transgene translation by quantifying WWOX protein levels at multiple postnatal time points. Immunoblot analyses at P30 revealed a marked dose-dependent increase in WWOX protein, most prominently in the cortex and to a lesser extent in the hippocampus, midbrain, and cerebellum, with a trend toward higher expression in HD-treated mice compared with LD-treated mice (Figures 5I–5L). At later stages (∼3 months), both immunoblot and immunohistochemistry analyses demonstrate that HD-treated mice maintained elevated WWOX protein levels across all examined brain regions (Figures S5A- S5D, S5H, and S5I), a pattern also evident in the spinal cord (Figure S6A). Notably, mice from either treatment group that failed to survive exhibited reduced WWOX expression, reinforcing the link between effective protein restoration and survival (Figures S5A–S5D). Long-term analyses demonstrated sustained WWOX expression. Robust WWOX protein levels persisted at P180 (Figures S5E–S5G), and widespread, stable expression was maintained at P240 and P300 throughout the cortex, hippocampus, midbrain, cerebellum, and spinal cord following a single neonatal injection (Figures S5J–S5K; Figures S6B–S6D). Consistent with physiological expression, WWOX protein was also detected in the sciatic nerve of HD-treated mice (Figures S6E–S6G), thus supporting functional relevance in the peripheral nervous system. In contrast, no WWOX expression was detected in the liver following either LD or HD treatment (Figures S6H and S6I), indicating that transgene expression was restricted to the central and peripheral nervous systems without detectable peripheral leakage. Together, these results demonstrate that an optimized AAV9-hSynI-WWOX gene therapy enables robust, dose-dependent, neuron-restricted, and sustained restoration of WWOX DNA, mRNA, and protein expression across the nervous system, supporting its durable therapeutic efficacy. Loss of WWOX disrupts myelination, while neuronal WWOX restoration promotes dose-dependent myelination and suppresses neuroinflammation Efficient neuronal signaling depends on proper myelination, a postnatal process that is tightly coupled to neuronal activity. Accumulating evidence indicates that excitatory neurons regulate oligodendrocyte precursor cell (OPC) proliferation and myelin biosynthesis through activity-dependent, non-cell autonomous mechanisms, although the genetic mediators of neuron-oligodendrocyte communication remain incompletely defined.50,51,52,53,54,55,56,57,58,59,60,61,62 Recent studies have implicated WWOX in myelin-related pathologies, including Alzheimer’s disease and multiple sclerosis, where its loss exacerbates neurodegeneration and demyelination.24,25,31 To assess the impact of WWOX deficiency and neuronal WWOX restoration on myelination, we performed MBP immunostaining on coronal brain sections from WT, Wwox-null, and AAV9-hSynI-hWWOX-treated mice. Consistent with this emerging link, Wwox-null mice exhibited profound hypomyelination accompanied by impaired neuronal activity across multiple brain regions, including the CC, cortex, striatum, and anterior commissure (AC) (Figures 6A–6E). Strikingly, neuron-specific re-expression of WWOX using AAV9-hSynI-hWWOX resulted in a robust, dose-dependent restoration of myelination. HD treatment achieved near-complete rescue across affected regions, whereas LD treatment produced only partial recovery relative to HD (Figures 6F; S7I). These data suggest neuronal WWOX as a critical upstream regulator of postnatal myelin development and maintenance. Neuronal WWOX restoration rescues hypomyelination in Wwox-null mice in a dose-dependent manner (A) Representative coronal brain sections stained for myelin basic protein (MBP) showing reduced myelination in KO mice compared with WT controls at P18. (B) MBP immunohistochemistry staining in the corpus callosum (CC), cortex, striatum, and anterior commissure (AC) of WT and KO mice at P18. (C–E) Bar graphs showing significant reduction in brain weight (C), CC thickness (D), and MBP intensity across regions (E) in KO mice relative to WT controls. Percentages (E) indicate relative MBP intensity compared with WT. (F) Representative MBP staining in WT and AAV9-hSynI-hWWOX-treated KO mice at LD (1.23 × 1011 vg) and HD (2.63 × 1011 vg), demonstrating dose-dependent restoration of myelination across CC, striatum, and cortex. Statistical analysis was performed using Student’s t test (∗∗p < 0.01, ∗∗∗p < 0.001). Error bars represent mean ± SD. In parallel, Wwox-null mice displayed pronounced astrogliosis and microglial activation, consistent with a heightened neuroinflammatory state (Figures S7A–S7D). Neuronal WWOX restoration significantly reduced both astrocyte reactivity and microglial density in a dose-responsive manner, indicating effective resolution of neuroinflammation (Figures S7E–S7H). Together, these findings demonstrate that WWOX loss disrupts myelin markers and induces gliosis, and neuron-targeted restoration normalizes MBP and glial reactivity in a dose-dependent manner. This dual effect underscores the therapeutic potential of neuronal WWOX gene therapy for WWOX-related neurodevelopmental and neurodegenerative disorders. WWOX deficiency promotes early epileptiform activity, while optimized AAV-mediated WWOX delivery suppresses spike-wave discharges Based on our previous studies demonstrating that both Wwox-null and neuron-specific Wwox KO mice exhibit pronounced epileptiform activity,43 we sought to directly assess whether HD WWOX gene therapy modulates early network hyperexcitability. To this end, we implemented continuous electrocorticography (ECoG) recordings to quantify epileptiform activity during early postnatal development. Continuous ECoG monitoring was performed in WT and Wwox-null pups beginning at approximately P14 after ICV administration at P0 of RI or AAV9-hSynI-hWWOX, respectively. Representative ECoG traces from KO pups revealed frequent interictal-like discharges and spike events, whereas WT pups displayed relatively stable baseline activity with sparse spiking (Figures 7A and 7B). Quantitative analysis of daily spike rates over a 7-day recording period demonstrated a consistently higher incidence of epileptiform spikes in KO pups compared with WT controls (Figure 7B). Averaged spike counts further confirmed a significant elevation in spike activity in KO animals (Figure 7C), indicating that WWOX deficiency leads to early-onset neuronal hyperexcitability during postnatal development. Wwox KO littermates exhibit elevated spike activity compared with WT pups, and AAV-mediated WWOX restoration rescues SWDs in Wwox KO pups (A) A single representative ECoG trace together with a magnified segment of pups at ∼P14 showing recurrent spikes. A zoomed-in segment highlights a single spike discharge. (B) Quantification of daily spike frequency (spikes/hour) over 7 days of recording. KO pups consistently displayed higher spike rates compared with WT littermates. (C) Averaged spike counts across the recording period in WT and KO pups. Data are shown as mean ± SEM; n = 5 littermates per group. Statistical significance was determined using Student’s t test. ∗p < 0.05; ns, non-significant. (D) Representative ECoG trace from a WWOX KO pup at ∼P14 showing a characteristic burst of generalized spike-wave discharges (SWDs). Magnified segment highlights rhythmic SWD activity typical of the KO group. (E) Quantification of SWD events per animal across groups (WT, KO, and KO + AAV9-hSynI-hWWOX [HD]). WWOX KO pups exhibited a significantly higher number of SWDs relative to WT, whereas WWOX-rescued pups (HD) showed a marked reduction in SWD incidence. Data are presented as mean ± SEM (n = 5 pups per group). Statistical significance was determined using Student’s t test. ∗∗∗∗p < 0.0001; ns, non-significant. We next assessed whether WWOX restoration could modulate this epileptiform phenotype. In addition to elevated spike activity, KO pups exhibited prominent spike-wave discharges (SWDs), characterized by rhythmic, high-amplitude spike-wave complexes, which were largely absent in WT animals (Figure 7D). Notably, neuronal rescue of WWOX in KO pups using HD AAV-mediated gene delivery markedly reduced SWD incidence, restoring network activity toward WT levels (Figure 7E). Together, these analyses demonstrate that WWOX loss increases early developmental epileptiform activity, whereas AAV-mediated WWOX restoration effectively suppresses pathological SWDs. Although based on a limited cohort, these findings support a direct role for WWOX in regulating early neuronal network excitability and highlight the therapeutic potential of WWOX gene therapy in mitigating epilepsy-related phenotypes. Early postnatal WWOX gene therapy achieves durable therapeutic rescue All preceding treatments were administered at P0‑P1. To determine whether later ICV delivery could still confer therapeutic benefit, we performed ICV injection of AAV9-hSynI-hWWOX in Wwox-null mice at daily intervals from P0 to P5, with at least three littermates treated per time point. WT littermates received identical injections to control for procedural effects. This developmental period corresponds to rapid brain growth and circuit formation in mice and requires age-specific stereotaxic targeting. Strikingly, neuronal WWOX restoration using the high-dose vector at any time point between P0 and P5 was sufficient to fully rescue the Wwox-null phenotype (Figure S8). Treated mice exhibited a dramatic extension in survival, from approximately three weeks to nearly one year, along with normalization of body weight and blood glucose levels (Figures S8A–S8D). Moreover, mice showed improved motor coordination and restored fertility (data not shown). Consistently, immunoblot and immunohistochemistry analyses confirmed robust WWOX expression with widespread distribution across major brain regions, including the cortex, hippocampus, and cerebellum (Figures S8E and S8F). Early WWOX re-expression also markedly enhanced myelination and reduced astrogliosis (Figures S8G–S8I). Collectively, these results demonstrate that timely neuronal re-expression of WWOX during early postnatal development is sufficient to reverse the otherwise lethal neurological and glycemia defects of Wwox deficiency. Therapeutic rescue beyond this early postnatal window was not explored, as Wwox-null mice rapidly deteriorate with progressive neurological dysfunction and early lethality, precluding effective intervention at later stages. Discussion In this study, we establish that optimized AAV9-mediated neuronal replacement of WWOX provides a robust and translationally relevant strategy that rescues the full spectrum of neurological and systemic abnormalities in Wwox-null mice. By systematically interrogating promoter choice, cell-type specificity, and expression strength, dose, and timing, we define the key parameters governing therapeutic efficacy. Our results establish that neuronal WWOX expression is both necessary and sufficient for survival and CNS homeostasis, whereas expression in oligodendrocytes alone is insufficient to confer benefit. Gene therapy has demonstrated meaningful clinical success, fundamentally transforming the outlook for patients with previously untreatable genetic disorders. A growing body of work supports AAV-mediated gene replacement as a viable therapeutic strategy for monogenic neurodevelopmental disorders when appropriate cellular targeting and expression control are achieved. Neuron-directed AAV delivery has been shown to rescue structural, electrophysiological, and behavioral phenotypes in mouse models of SCN1A-63 and SCN1B-64associated Dravet syndrome, SLC6A1 deficiency,65 STXBP1-related disorders,66 and FOXG1 syndrome,67 including correction of severe neuroanatomical defects such as CC agenesis. In parallel, multicenter preclinical studies of MECP2 gene therapy in Rett syndrome have demonstrated robust efficacy across models and favorable safety profiles in nonhuman primates.68 Collectively, these advances provide compelling evidence that targeted gene delivery to the nervous system can achieve durable gene expression, modify underlying disease biology, and produce clinically meaningful benefits, particularly when intervention occurs early during critical windows of brain development. In our study, under neonatal AAV9 ICV delivery conditions, neuronal WWOX expression driven by the hSynI promoter produced the most robust and sustained therapeutic benefit, whereas ubiquitous or oligodendrocyte-restricted expression resulted in limited or transient rescue. Strong ubiquitous promoters (CMV and CBA) supported early postnatal improvement in survival and glycemic control, indicating that high-level early expression can transiently ameliorate disease manifestations; however, these effects were not maintained at later stages. In contrast, MBP-driven expression failed to confer measurable benefit, which may reflect the limited oligodendrocyte tropism of AAV9 following neonatal ICV administration rather than a lack of relevance for oligodendrocyte WWOX expression. We also cannot exclude that the optimal dose was not achieved, as all vectors were tested at the same titer (4E10). Notably, a limitation of our study is that WWOX expression levels were not normalized across promoter conditions, and vector genome copy number, transcript abundance, or protein levels were not systematically quantified across all constructs. Thus, differences in therapeutic durability could reflect a combination of cell-type specificity, expression magnitude, and temporal kinetics. Nonetheless, the observed promoter-dependent outcomes are consistent with our prior conditional KO studies, in which deletion of Wwox in neural stem/progenitor cells (Nestin-Cre) or postmitotic neurons (Synapsin I-Cre) recapitulated the severe neurological and metabolic phenotypes of global KO mice, whereas astrocyte- (GFAP-Cre) or oligodendrocyte-specific (Olig2-Cre) deletion produced no overt abnormalities.42 Taken together, under matched neonatal AAV9-ICV conditions, neuron-restricted WWOX expression emerged as the most effective and durable strategy, while ubiquitous and oligodendrocyte-restricted approaches showed limited durability within the parameters tested. Another central insight from the present study is the importance of precise control over transgene expression for both efficacy and safety. Our earlier therapeutic construct incorporated the WPRE element, which is widely used to enhance transgene expression and vector yield. Consistent with this function, inclusion of WPRE markedly increased WWOX protein levels in a dose-dependent and regionally consistent manner across multiple brain areas. However, WPRE-containing vectors also produced a more intense and spatially diffuse neuronal signal, characterized by prominent punctate staining throughout neurons, likely reflecting elevated expression levels rather than altered subcellular targeting. Such amplification may increase peak transgene exposure and introduce variability across developmental stages and brain regions. Although no overt toxicity was observed in prior studies, we removed WPRE as a proactive risk-mitigation step to improve the predictability and control of neuronal WWOX expression for potential clinical translation. In the absence of WPRE, WWOX expression was reduced but exhibited a more uniform, predominantly cytoplasmic distribution that closely resembled endogenous WWOX patterns in WT neurons. We subsequently compensated for lower expression through dose optimization, thereby preserving therapeutic efficacy while maintaining a tighter and more controlled expression envelope. Using this optimized vector configuration, we identified a clear dose-response relationship, providing strong mechanistic evidence of on-target activity and enabling definition of an optimal therapeutic dose. Neuron-specific WWOX expression driven by the synapsin I promoter resulted in graded improvement across seizures, ataxia, hypoglycemia, infertility, hypomyelination, neuroinflammation, and premature mortality, with optimal dosing achieving near-complete phenotypic normalization even in the absence of WPRE. Importantly, electrophysiological analyses revealed that Wwox-null pups display early-onset neuronal hyperexcitability, characterized by frequent interictal-like spikes and SWDs, which are hallmarks of epileptic encephalopathy. AAV-mediated neuronal restoration of WWOX robustly suppressed these pathological discharges, supporting a direct role for neuronal WWOX in maintaining network stability during early brain development (Figure 7). Although conducted in a limited cohort, these findings support the capacity of WWOX gene therapy to correct epileptiform activity at its developmental onset. An additional key finding is the identification of an early postnatal period during which WWOX restoration is tractable in this murine model. Consistent with the experience from other gene therapy approaches targeting severe neurodevelopmental disorders,63,64,65,66,67,68 neonatal or very early postnatal intervention was required to achieve durable rescue. Delivery of AAV9-hSynI-hWWOX within the P0‑P5 window normalized growth, markedly extended survival to nearly one year, restored motor and glycemic function, and prevented progressive neurodegeneration (Figure S8). Therapeutic intervention beyond this period was not feasible, as Wwox-null mice rapidly deteriorate after P5, developing severe growth retardation, progressive weight loss, and frequent seizures.42,69 These features rendered the animals highly vulnerable and raised ethical and practical concerns regarding invasive procedures such as ICV injection at later stages. Accordingly, the inability to assess later intervention likely reflects a combination of model-specific biological constraints and technical limitations, rather than a definitive boundary for therapeutic responsiveness. These findings should, therefore, be interpreted in the context of murine neurodevelopment, where rapid postnatal brain maturation and accelerated disease progression may impose a narrower therapeutic window than in humans. In contrast, human patients with WWOX-related encephalopathies may exhibit different developmental trajectories, disease kinetics, and therapeutic responsiveness, potentially allowing for later intervention. Accordingly, future studies should explore alternative strategies to extend the therapeutic window, including earlier prenatal delivery, less invasive or systemic administration routes, combinatorial approaches to stabilize neuronal networks, or adjunct therapies aimed at mitigating seizures, inflammation, or metabolic stress prior to gene replacement. Together, these considerations highlight both the strengths and limitations of the murine model and underscore the need for continued optimization to translate WWOX gene therapy effectively to human patients. In summary, our study defines a comprehensive therapeutic framework for WWOX-related encephalopathies based on neuronal specificity, dose optimization, and developmental timing. Neuron-targeted AAV9-mediated WWOX replacement achieves durable rescue of both neurological and systemic phenotypes in Wwox-null mice, establishing a strong preclinical rationale for translation. Given the established safety profile of AAV9 in pediatric applications, the neuronal precision of the Synapsin I promoter, and the feasibility of neonatal ICV delivery, this approach represents a viable path toward clinical intervention in infants with WWOX deficiency. Collectively, these findings position WWOX as a central regulator of neuronal homeostasis and identify its timely restoration as a promising therapeutic strategy for otherwise fatal developmental encephalopathies. Materials and methods Plasmid vectors hWWOX cDNA was cloned under different promoters, including hSynI, EF1-alpha, CBA, MBP, and CMV, in the pAAV backbone. Constructs driven by EF1α, CMV, and MBP included WPRE, whereas the hSynI-driven vector was generated both with and without WPRE. These vectors were packaged into AAV9 serotypes (Fujifilm Diosynth Biotechnologies, Watertown, MA, USA; Vector Biolabs, Philadelphia, PA, USA; Boston Institute of Biotechnology [BIB], Boston, MA, USA). Custom-made AAV9-CBA-hWWOX and AAV9-hSynI-EGFP viral particles were obtained from the Vector ELSC Core Facility at the Hebrew University of Jerusalem. Viral titers were determined by RT-qPCR using bGH primers. Mice The generation of Wwox-null (−/−) mice (KO) was previously reported.69 Mice were kept on an FVB (Friend leukemia Virus B) background. Heterozygote (+/−) or KO rescued mice (KO injected with AAV9-hSynI-hWWOX) were used for breeding to generate KO mice. After weaning, WT and KO rescued mice were separated by sex and housed in individual ventilated cages (IVC). To assess fertility, one male and one female of the same genotype were housed together in an IVC cage. All animals were maintained in a specific pathogen-free (SPF) facility under a 12-h light/dark cycle, with ad libitum access to food and water. All animal experiments were conducted in accordance with the ethical guidelines and with prior approval from the Hebrew University Institutional Animal Care and Use Committee (HU-IACUC). Genotyping Genomic DNA was extracted from the tail/ear using DNA lysis buffer (20 mM NaOH, 0.1 mM EDTA). PCR reaction was performed using Taq DNA polymerase (2× rapid Taq master mix-P222-01-Vazyme) and the following primers were used: shared 5′ primer of the WT Wwox allele (WF, 5′ GCAGAATGTCTTGCTAGAGCTTTG 3′), 3′ primer in sequence deleted in the targeted allele (WR, 5′ ATACTGACATCTGCCTCTAC 3′), and a 3′ primer internal to the Geo cassette (GeoR1, 5′ CAAAAGGGTCTTTGAGCACCAGAG 3′). ICV injection of AAV particles into P0-P5 Wwox-null mice ICV injections of either AAV9-hWWOX (under different promoters) or reference item (RI; PBS with 5% sorbitol and 0.001% pluronic F-68) were conducted using stereotactic technique to ensure consistency by following a published protocol.70 Briefly, Wwox-null (KO) or WT neonates were anesthetized by placing them on a dry, flat, cold surface. The anesthetized pup head was gently wiped with a cotton swab soaked in 70% ethanol. The pup (KO or WT) was staged on the stereotaxic frame for AAV9-hWWOX or RI delivery. The injection site was targeted at x = ±0.8 mm, y = 1.5 mm, and z = −1.6 mm, relative to lambda. These anatomical landmarks were visible through the skin at P0‑P1. For older pups, the coordinates were adjusted to account for developmental changes; for P5 pups, the injection site was targeted at x = ±1.0 mm, y = 1.0 mm, and z = −2.0 mm. A Micro-4 nano-pump controller was used to ensure a steady injection rate of 1–1.5 μL/min, delivering 2.0 μL/hemisphere through a Hamilton syringe with a 32G needle (World Precision Instruments). The needle was kept at the injection site for 30‑60 s to allow proper diffusion and then removed slowly over 1 min. The procedure was repeated for the contralateral hemisphere. After the injection, the pups were placed on a warming pad till they awoke and then returned to their mother’s cage. The pups were closely monitored for growth rate, glucose levels, seizures, ataxia, and general condition to assess phenotypes. Blood glucose levels Blood glucose was measured at P10, P20, and P30 from the tail nick using a glucometer (Roche Diagnostics, Mannheim, Germany). DNA and RNA extraction Brain tissue samples were collected, snap-frozen in liquid nitrogen, and preserved in −80°C for analysis. DNA extraction was performed using the DNeasy Blood and Tissue Kit (Qiagen) according to the manufacturer’s protocol. Briefly, tissue samples (up to 25 mg) were lysed in ATL buffer supplemented with Proteinase K at 56°C overnight. Following lysis, samples were mixed with AL buffer and 100% ethanol to facilitate DNA binding to the DNeasy Mini Spin Column. The bound DNA was washed sequentially with AW1 and AW2 buffers to remove contaminants and ensure that only pure DNA remains bound to the column. Next, AE buffer was added to the mini spin column and incubated for 5 min at room temperature (RT) for elution of purified DNA. DNA integrity and concentration were assessed prior to downstream applications by using DeNovix (DS-11FX+). For qPCR, 50 ng of DNA template was used per reaction. The primer sequences were as follows: forward 5′ GCTCTCTTAAGGTAGCCCCG 3′, reverse 5′ CGCCTCATCCTGGTCCTAAA 3′. Total RNA was isolated from non-perfused tissue using TRIzol reagent (Bio-Tri) according to the manufacturer’s phenol-chloroform protocol. Briefly, tissues were homogenized in TRIzol at 1 mL per 50–100 mg tissue, followed by phase separation with chloroform. The aqueous phase was recovered, RNA was precipitated with isopropanol, and the pellet was washed with 75% ethanol, air-dried, and resuspended in DEPC-treated nuclease-free water. RNA concentration and integrity were assessed on a DeNovix (S-11 FX+) spectrophotometer. Extracted RNA was stored at −80°C until use. cDNA was synthesized from 1 μg total RNA using the qScript cDNA Synthesis Kit (Quantabio) following the manufacturer’s instructions. For qPCR, 50 ng of cDNA template was used per reaction. The primer sequences were as follows: forward, 5′ ACCTACTTGGACCCAAGACTGGCG 3′, reverse, 5′ GGTGCTGCCGTCGTATCTTTGCC 3′. Immunoblot Brain tissue samples were homogenized using protein lysis buffer (50 mM Tris, pH 7.5, 150 mM NaCL, 10% glycerol, 0.5% Nonidet P-40 [NP-40]) with the addition of protease and phosphatase inhibitors. The lysates were subjected to SDS-PAGE under standard conditions. The following antibodies were used: rabbit polyclonal anti-WWOX 1:10,000; mouse monoclonal anti-GAPDH (CB-1001), Calbiochem 1:10,000; rabbit polyclonal anti-HSP90 (4874 S), cell signaling 1:1000. Immunohistochemistry Mice (WT, KO, and KO injected mice) at different ages (P10‑P180) were deeply anesthetized using an overdose of ketamine and xylazine and then transcardially perfused using 4% paraformaldehyde (PFA)/PBS. Dissected brains were post-fixed in 4% PFA overnight at 4°C, followed by incubation in 30% sucrose overnight at 4°C, and were then embedded in OCT. Sagittal brain sections (14 μm) were prepared using a cryostat and stored at −80°C. Sections were warmed to RT, washed with PBS, and fixed with 4% PFA for 5 min, followed by PBS washing. Permeabilization was done using PBT (0.1% Triton X-100 in PBS). Formalin-fixed, paraffin-embedded (FFPE) sagittal tissue sections (14 μm) were deparaffinized, followed by antigen retrieval with 25 mM citrate buffer (pH = 6) in a pressure cooker at high pressure for 3 min. The sections were then left to cool for 25 min. Fresh coronal sections were prepared using a vibratome (50 μm). The sections (FFPE, frozen, fresh sections) were then washed and blocked using a blocking buffer (5% normal goat serum, 0.5% BSA and 0.1 %Triton X-100 in PBT) followed by incubation with a specific primary antibody in a humidified chamber overnight (rabbit polyclonal anti-WWOX 1:5,000, mouse anti-beta III tubulin (B364854), biolegend 1:1000; rabbit anti-GFAP-clone EPR1034Y (AB68428), Abcam 1:250; rabbit anti-Iba1-clone HL22 (ab28937), Abcam 1:250; rabbit anti-MBP, (Ab65988), Abcam 1:200; and mouse anti-NeuN, (MAB377), Milipore, 1:500). The sections were washed using PBST buffer (0.05% Tween 20 in PBS) and incubated with the corresponding secondary fluorescent antibody for 1 h at RT (goat anti-rabbit Alexa flour 647 (Ab150079), Abcam 1:1000 and goat anti-mouse Alexa flour 488 (ab150117), Abcam 1:1000) with Hoechst 33258 (1:1000). Then, the sections were washed with PBST three times and were mounted using fluorescent mounting media (Dako s3023). Sections were then imaged using a 3DHISTECH panoramic scanner (3DHISTECH, Hungary). Sciatic nerve and spinal cord extraction Sciatic nerve extraction was performed following a previously published protocol.71 Briefly, in a transcardially perfused mouse, a dorsal incision along the hind leg was made to expose the sciatic nerve. Using fine-point tweezers and surgical scissors, the sciatic nerve was carefully dissected from its surrounding tissues. The nerve was handled delicately to avoid damage. For the spinal cord extraction, a midline incision along the dorsal part of the mouse was made with a scalpel. Then, the skin, underlying muscles, and bones were carefully removed to expose the spinal cord. Using forceps, the spinal cord was gently extracted from the vertebral column. Immediately after excision, the specimen (sciatic nerve and spinal cord) was either snap frozen for further analysis or immersed in 4% formalin to ensure rapid and uniform fixation for 24–48 h at RT. After fixation, the tissue was transferred to 70% ethanol for 24 h prior to processing in an automated tissue processor (Leica, Histocore). Then, the tissue was embedded in paraffin blocks, and sections were prepared using a microtome (14 μm). Behavioral tests Behavioral assessment was conducted on WT + RI and KO rescued mice (HD) at the age of 3 months (both males and females), as previously published.45 Animals were assessed in an open-field test, in an elevated plus maze, and on a rotarod. The open-field test was conducted following a previously published protocol.72 Briefly, mice were placed in a 50 cm × 50 cm × 33 cm arena and allowed to explore freely for 6 min. The center of the arena was defined as a 25 cm × 25 cm square. Key measurements, including the mice velocity, time spent in the central area, and time spent near the arena’s perimeter, were recorded. The test was monitored using a video camera connected to a computer with tracking software (Ethovision 12). The elevated plus maze consisted of two open arms (30 cm × 5 cm) with a 1 cm high rim and two closed arms (30 cm × 5 cm) with a 16 cm high rim, positioned perpendicular to each other. The maze was elevated 75 cm above the floor. Mice were placed into the maze and allowed to explore freely for 6 min. The time spent in both the open and closed arms was recorded to assess anxiety-like behaviors. For motor coordination and balance assessment, a rotarod test was performed. Each mouse was placed on a rotating rod, which gradually accelerated from 5 revolutions per minute (rpm) to 40 rpm over 99 s. Mice underwent three trials, with a 20-min interval between trials. The latency to fall off the rod was recorded, and if the mouse remained on the rod for 240 s without falling, the trial was terminated. Results from all trials were used to calculate the average latency to fall. DRG culture Neurons were cultured following a previously published protocol.45 Briefly, DRG (Dorsal Root Ganglion) neurons were isolated from mouse embryos at E13.5. Embryos were genotyped and the DRGs were collected in cold L-15 medium. Tissues were dissociated in 0.25% trypsin, triturated, centrifuged, and re-suspended in NB medium (Neurobasal, B27 supplement, 0.5 mM L-glutamine, and penicillin-streptomycin). Pre-cleaned 13 mm diameter glass coverslips were placed in 4-well dishes and coated with Matrigel (1 h at RT) then poly-D-lysine (30 min at RT) prior to dissection. Cells were plated at a density of 40,000 cells/13-mm coverslips in NB medium and maintained in a humidified incubator at 37°C and 5% CO2. Cultures were treated with fluorodeoxyuridine at days in vitro (DIV)2, DIV4, and DIV6 to eliminate non-neuronal cells. Fifty percent of the cell culture media was replaced every third day. At DIV6, Wwox-null DRG neurons were infected with AAV9-SynI-EGFP or AAV9-SynI/CMV/MBP-hWWOX and cultured for 3 days, and then, expression of GFP and WWOX were validated by immunohistochemistry. Surgery and ECoG data acquisition This study utilized WT mice, WWOX KO mice injected with RI, and WWOX-KO mice injected with AAV9-hSynI-hWWOX (rescue group). At P14, pups were anesthetized with isoflurane (1%–2%, adjusted for low body weight; Terrell, USP, Piramal Critical Care) and positioned in a stereotaxic frame (Kopf Instruments, CA, USA). Carprofen was administered prior to surgery, and animals were implanted with a single-channel ECoG transmitter (model A3049F2; Open-Source Instruments, USA). Unlike standard adult procedures, the transmitter body was kept outside of the body to accommodate the small size of pups. Two intracranial electrodes were placed epidurally: the recording electrode was positioned above the right dorsal cortex, and the reference electrode was placed contralaterally. Electrodes were secured to the skull with miniature screws and tissue adhesive, and the incision was sealed with dental cement. Immediately after surgery, continuous wireless ECoG acquisition was initiated without a recovery interval. Recordings were maintained for 7 consecutive days, during which pups were housed with their mother and monitored under continuous 24/7 ECoG activity. Signals were acquired and stored using Neuroarchiver software (Open-Source Instruments Inc.). ECoG traces were manually screened for interictal spikes and SWDs by a blinded investigator. Events were defined based on amplitude, duration, and morphology criteria established in previous mouse epilepsy models. A subset of events was cross-validated with time-synchronized video recordings to confirm electrographic abnormalities. Statistical analysis Statistical analyses were performed using GraphPad Prism (version 5 or later) and Microsoft Excel. Data were analyzed in a blinded manner when feasible and are presented as mean ± standard deviation (SD), as indicated in the figure legends. Comparisons between two groups were performed using two-tailed unpaired Student’s t tests. Survival was analyzed using Kaplan-Meier curves with significance assessed by the log-rank (Mantel-Cox) test. Longitudinal measurements (e.g., body weight and blood glucose) were analyzed at individual time points as specified in the figures. ECoG, behavioral, molecular, and histological quantifications were performed on per-animal averaged values and compared using unpaired t tests. Exact sample sizes (n) and p values are reported in the figure legends. A p value <0.05 was considered statistically significant; non-significant results are denoted as ns. Data and code availability Not relevant. Acknowledgments We thank all members of the Aqeilan laboratory and the Mahzi team for their contributions and technical support. We are especially grateful to the individuals affected by WOREE and SCAR12 syndromes and their families for their continued engagement and generosity. We especially thank Ms. Takwa Jebara (a former student at the Aqeilan lab), Dr. Yoni Kupchik, Dr. Kineret Inbar, Ms. Naheel Lawabny, Mr. Wajeeh Salaymeh, and Ms. Marah Bakhtan for technical assistance. This work was supported by Mahzi Therapeutics and the WWOX Foundation. Author contributions M.O., R.A., A.B., T.B., Y.W., and R.I.A. conceived the study; M.O. led the design and implementation of ICV gene therapy ; M.O., R.A., S.R., and B.A. designed and performed the in vivo experiments, including phenotyping, molecular analyses, histology/immunostaining, imaging, and behavioral testing; P.K.S. and T.S-A. performed ECoG experiments and analysis; M.O., R.A., and R.I.A. wrote the initial manuscript. All authors reviewed and edited the final manuscript. Declaration of interests R.I.A. is a consultant for Mahzi Therapeutics. A.B., T.B., and Y.W. are employed by Mahzi Therapeutics. Footnotes Supplemental information can be found online at https://doi.org/10.1016/j.omta.2026.201791. Supplemental information Document S1. Figures S1–S8 Document S2. Article plus supplemental information References Associated Data Supplementary Materials Document S1. Figures S1–S8 Document S2. Article plus supplemental information Data Availability Statement Not relevant.